The flame recognition system based on video image is applied to the fire flame detection, and it can be very good to avoid the shortage of the traditional fire flame detection system. Three frame difference algorithm is adopted to extract the motion features of the flame. And Lab color space is applied in analyzing the flame color. By analyzing the circular degree of the flame and the interference sources, the change rate of the flame area and the angle feature of the fire flame, it can be determined whether there is a fire in the image. The experimental results shows: this algorithm can not only effectively improve the accuracy and real-time performance of the flame detection and identification, but also eliminate the interference of other substances and reduce the misjudgment.

Aiming at the random delays and packet dropouts in ZigBee network, a Generalized Predictive Control (GPC) algorithm is proposed. The control sequence combine with timestamp, feedback sequence and queuing mechanisms to design a predictive control strategy. The GPC algorithm derivation, system design and implementation method are given, and then the algorithm is validated in a networked control system simulation toolbox based on TrueTime. The proposed algorithm shows better performance than conventional PID algorithm for a DC servo motor in control system based on ZigBee technology.

To address the problem that the dimension of the feature vector extracted by Local Binary Pattern (LBP) for face recognition is too high and Principal Component Analysis (PCA) extract features are not the best classification features, an efficient feature extraction method using LBP, PCA and Maximum scatter difference (MSD) has been introduced in this paper. The original face image is firstly divided into sub-images, then the LBP operator is applied to extract the histogram feature. and the feature dimensions are further reduced by using PCA. Finally,MSD is performed on the reduced PCA-based feature.The experimental results on ORL and Yale database demonstrate that the proposed method can classify more effectively and can get higher recognition rate than the traditional recognition methods.

Based on the knowledge of graph theory, this thesis builds mathematical model. With the concept of effective path and designed time, the method of MADM and improved Dijkstra, the paper solves the path evaluation problem about singular attribute, length and time of path, and both of the attributes.

Although the chord height error might be ensured, the adaptive algorithm of the NURBS curve circular interpolation feed rate is still likely to result in the speed fluctuation which is not in accordance with given acceleration or deceleration rules. On the foundation of the analytic principle of the NURBS curve circular interpolation, a kind of interpolating algorithm based on the NURBS curve continuous small segments of S-shaped acceleration and deceleration is presented. And considering the chord height error, the variable step and feed rate can be effectively combined to make a better estimation. The simulation results on real examples show that the method not only simplifies the complicated calculation of curve interpolation process, simultaneously improves the speed of Motion Smoothing Implementation for NURBS curve interpolation, but also ensures the interpolation precision. The algorithm can be commonly applied in real manufacturing for high-speed and high-precision curve process.